About

Yasamin Mostofi is a Professor in the Department of Electrical and Computer Engineering at UC Santa Barbara. Her research interests include RF Sensing and Learning, Communication-Aware Robotics, Next-generation Communication, Vision, and Robotics. She is associated with the RF Sensing and Robotic Network Lab. Her contact information includes a phone number (+1 805-893-4251), email (ymostofi@ece.ucsb.edu), and her office is located in Harold Frank Hall, Room 5121. She is also affiliated with the Visiting Department and the RF Sensing and Robotic Network Lab, contributing to advancements in wireless sensing, robotic communication, and integrated systems.

Research topics

• Computer Science
• Artificial Intelligence
• Engineering
• Telecommunications
• Electrical engineering
• Control engineering
• Physical medicine and rehabilitation
• Database
• Distributed computing
• Real-time computing
• Computer network

Selected publications

• mmRidge: Revealing Emergent Crowd Flow Structures with mmWave MIMO Radar

  2026-02-25

  articleOpen accessSenior author

  In this paper, we propose mmRidge, a novel framework that reveals emergent crowd flow structures using commodity mmWave MIMO radar. More specifically, mmRidge reveals flow boundaries that separate distinct motion behaviors in dense crowds. Our approach first converts radar point clouds into high-fidelity 2D flow fields that capture the collective motion of people. Building on concepts from dynamical systems theory, mmRidge computes a scalar Finite-Time Lyapunov Exponent (FTLE) field to uncover Lagrangian Coherent Structures (LCS) as “ridges” that demarcate boundaries between distinct crowd motion behaviors. Experimental validation with crowds of 18 or more individuals, across three real-world environments, demonstrates the effectiveness of mmRidge in detecting flow separators, highlighting the impact of advection, time horizons, and directionality on structure identification. Overall, the paper shows that mmWave radar can provide a low-power, privacy-preserving modality for identifying spatial crowd flow structures in dense and complex environments.

  PublisherDOI

• mmFlux: Crowd Flow Analytics with Commodity mmWave MIMO Radar

  ArXiv.org · 2025-07-09

  preprintOpen accessSenior author

  In this paper, we present mmFlux: a novel framework for extracting underlying crowd motion patterns and inferring crowd semantics using mmWave radar. First, our proposed signal processing pipeline combines optical flow estimation concepts from vision with novel statistical and morphological noise filtering. This approach generates high-fidelity mmWave flow fields-compact 2D vector representations of crowd motion. We then introduce a novel approach that transforms these fields into directed geometric graphs. In these graphs, edges capture dominant flow currents, vertices mark crowd splitting or merging, and flow distribution is quantified across edges. Finally, we show that analyzing the local Jacobian and computing the corresponding curl and divergence enables extraction of key crowd semantics for both structured and diffused crowds. We conduct 21 experiments on crowds of up to 20 people across 3 areas, using commodity mmWave radar. Our framework achieves high-fidelity graph reconstruction of the underlying flow structure, even for complex crowd patterns, demonstrating strong spatial alignment and precise quantitative characterization of flow split ratios. Finally, our curl and divergence analysis accurately infers key crowd semantics, e.g., abrupt turns, boundaries where flow directions shift, dispersions, and gatherings. Overall, these findings validate mmFlux, underscoring its potential for various crowd analytics applications.

  PublisherOA PDFDOI

• mmSnap: Bayesian One-Shot Fusion in a Self-Calibrated mmWave Radar Network

  ArXiv.org · 2025-05-01

  preprintOpen access

  We present mmSnap, a collaborative RF sensing framework using multiple radar nodes, and demonstrate its feasibility and efficacy using commercially available mmWave MIMO radars. Collaborative fusion requires network calibration, or estimates of the relative poses (positions and orientations) of the sensors. We experimentally validate a self-calibration algorithm developed in our prior work, which estimates relative poses in closed form by least squares matching of target tracks within the common field of view (FoV). We then develop and demonstrate a Bayesian framework for one-shot fusion of measurements from multiple calibrated nodes, which yields instantaneous estimates of position and velocity vectors that match smoothed estimates from multi-frame tracking. Our experiments, conducted outdoors with two radar nodes tracking a moving human target, validate the core assumptions required to develop a broader set of capabilities for networked sensing with opportunistically deployed nodes.

  PublisherOA PDFDOI

• mmFlux: Crowd Flow Analytics with Commodity mmWave MIMO Radar

  npj Wireless Technology · 2025-12-03

  articleOpen accessSenior author

  Abstract In this paper, we present mmFlux: a novel framework for extracting underlying crowd motion patterns and inferring crowd semantics using mmWave radar. First, our proposed signal processing pipeline combines optical flow estimation concepts from vision with novel statistical and morphological noise filtering. This approach generates high-fidelity mmWave flow fields-compact 2D vector representations of crowd motion. We then introduce a novel approach that transforms these fields into directed geometric graphs. In these graphs, edges capture dominant flow currents, vertices mark crowd splitting or merging, and flow distribution is quantified across edges. Finally, we show that analyzing the local Jacobian and computing the corresponding curl and divergence enables the extraction of key crowd semantics for both structured and diffused crowds. We conduct 21 experiments on crowds of up to 20 people across 3 areas, using commodity mmWave radar. Our framework achieves high-fidelity graph reconstruction of the underlying flow structure, even for complex crowd patterns, demonstrating strong spatial alignment and precise quantitative characterization of flow split ratios. Finally, our curl and divergence analysis accurately infers key crowd semantics, e.g., abrupt turns, boundaries where flow directions shift, dispersions, and gatherings. Overall, these findings validate mmFlux, underscoring its potential for various crowd analytics applications.

  PublisherOA PDFDOI

• Embracing Diffraction: A Paradigm Shift in Wireless Sensing and Communication

  ArXiv.org · 2025-05-02

  preprintOpen accessSenior author

  Wireless signals are integral to modern society, enabling both communication and increasingly, environmental sensing. While various propagation models exist, ranging from empirical methods to full-wave simulations, the phenomenon of electromagnetic diffraction is often treated as a secondary effect or a correction factor. This paper positions diffraction as a fundamentally important and underutilized mechanism that is rich with information about the physical environment. Specifically, diffraction-inducing elements generate distinct signatures that are rich with information about their underlying properties such as their geometries. We then argue that by understanding and exploiting these relationships, diffraction can be harnessed strategically. We introduce a general optimization framework to formalize this concept, illustrating how diffraction can be leveraged for both inverse problems (sensing scene details such as object geometries from measured fields) and design problems (shaping radio frequency (RF) fields for communication objectives by configuring diffracting elements). Focusing primarily on edge diffraction and Keller's Geometrical Theory of Diffraction (GTD), we discuss specific applications in RF sensing for scene understanding and in communications for RF field programming, drawing upon recent work. Overall, this paper lays out a vision for systematically incorporating diffraction into the design and operation of future wireless systems, paving the way for enhanced sensing capabilities and more robust communication strategies.

  PublisherOA PDFDOI

• Gait Disorder Assessment Based on a Large-Scale Clinical Trial: WiFi vs. Video vs. Doctor's Visual Inspection

  ArXiv.org · 2025-02-07

  preprintOpen accessSenior author

  Neurological gait disorders affect a large population, significantly reducing life quality. This paper brings a foundational understanding to the potentials of emerging sensing modalities (e.g., WiFi) for gait disorder assessment, via conducting a one-year-long clinical trial in collaboration with the Neurology Associates of Santa Barbara. Our medical campaign encompasses 114 real subjects and a wide spectrum of disorders (e.g., Parkinson's, Neuropathy, Post Stroke, Dementia, Arthritis). We then develop the first WiFi-based gait disorder sensing system of its kind, distinguished by its scope of validation with a large and diverse patient cohort. To ensure generalizability, we mainly leverage publicly-accessible online videos of gait disorders for training, and develop a video-to-RF pipeline to convert them to synthetic RF training data. We then extensively test the system in a neurology center (i.e., the Neurology Associates of Santa Barbara). Additionally, we provide a 1-1 comparison with a vision-based system, by developing a vision-based gait assessment system under identical conditions, a first-of-its-kind comparison to our knowledge. We finally contrast both systems with neurologists' accuracy when basing evaluation solely on visual gait inspection, by designing/distributing a large survey to 70 neurologists, offering the first apples-to-apples comparison of these three sensing modalities. Our findings can help integrate these sensing systems into medical practice, working towards equitable healthcare.

  PublisherOA PDFDOI

• Uncrewed Vehicles in 6G Networks: A Unifying Treatment of Problems, Formulations, and Tools

  Proceedings of the IEEE · 2025-03-12 · 5 citations

  articleSenior author

  Uncrewed vehicles (UVs) functioning as autonomous agents are anticipated to play a crucial role in the sixth generation (6G) of wireless networks. Their seamless integration, cost-effectiveness, and additional controllability through motion planning make them an attractive deployment option for a wide range of applications, both as assets in the network e.g., mobile base stations (BSs) and as consumers of network services (e.g., autonomous delivery systems). However, despite their potential, the convergence of UVs and wireless systems brings forth numerous challenges that require attention from both academia and industry. This article then aims to offer a comprehensive overview, encompassing the transformative possibilities as well as the significant challenges associated with UV-assisted next-generation wireless communications. Considering the diverse landscape of possible application scenarios, problem formulations, and mathematical tools related to UV-assisted wireless systems, the underlying core theme of this article is the unification of the problem space, providing a structured framework to understand the use cases, problem formulations, and necessary mathematical tools. Overall, this article sets forth a clear understanding of how UVs can be integrated in the 6G ecosystem, paving the way toward harnessing the full potential at this intersection.

  PublisherDOI

• Relay Incentive Mechanisms Using Wireless Power Transfer in Non-Cooperative Networks

  IEEE Transactions on Wireless Communications · 2025-04-30 · 2 citations

  articleSenior author

  The advances of 6G systems have prompted the study of new communication paradigms, including relay networks enabled by wireless power transfer (WPT). While existing literature focuses on the cooperative case, this paper examines a non-cooperative scenario in which a source must incentivize one of several battery-powered user equipments (UEs) to act as a relay by offering payment in the form of WPT, while the utility-maximizing UEs seeking to extract as much energy from the source as possible. We propose a protocol based on a reverse auction that enables the source to determine which candidate UE to select as the relay and the amount of energy to be transferred as payment, even when the channel quality between the candidates and the destination is unknown to the source. We first examine the performance of the system under the classical Vickrey auction. We prove that our protocol achieves the best possible outage probability and point out ways to mitigate the gap in energy efficiency when compared to a cooperative baseline. We then analyze system performance under a Myerson auction, which maximizes the auctioneer’s utility. To ensure computational tractability, we extend the forward auction regularity condition to the reverse setting, providing a mathematical characterization of regularity and the associated pricing mechanism. We then prove the regularity of the WPT-based auction under both lognormal and Rayleigh fading. To validate our analytical findings, we present extensive numerical results demonstrating how system parameters affect energy efficiency and outage probability. Our results show that auction-based protocols can reduce both outage probability and communication energy by more than 50% with as few as two relay candidates, compared to direct transmission by the source. Additionally, they demonstrate exponential convergence to the cooperative lower performance bound and indicate conditions under which one auction type is preferred over the other. Overall, the auction-based system offers a foundation for improved energy efficiency and communication reliability in non-cooperative environments.

  PublisherDOI

• Fast and Robust Stationary Crowd Counting with Commodity WiFi

  ArXiv.org · 2025-07-18

  preprintOpen accessSenior author

  This paper introduces a novel method for estimating the size of seated crowds with commodity WiFi signals, by leveraging natural body fidgeting behaviors as a passive sensing cue. Departing from prior binary fidget representations, our approach leverages the bandwidth of the received signal as a finer-grained and robust indicator of crowd counts. More specifically, we propose a mathematical model that relates the probability density function (PDF) of the signal bandwidth to the crowd size, using a principled derivation based on the PDF of an individual's fidget-induced bandwidth. To characterize the individual fidgeting PDF, we use publicly available online videos, each of a seated individual, from which we extract body motion profiles using vision techniques, followed by a speed-to-bandwidth conversion inspired by Carson's Rule from analog FM radio design. Finally, to enhance robustness in real-world deployments where unrelated motions may occur nearby, we further introduce an anomaly detection module that filters out non-fidget movements. We validate our system through 42 experiments across two indoor environments with crowd sizes up to and including 13 people, achieving a mean absolute error of 1.04 and a normalized mean square error of 0.15, with an average convergence time of 51 seconds, significantly reducing the convergence time as compared to the state of the art. Additional simulation results demonstrate scalability to larger crowd sizes. Overall, our results show that our pipeline enables fast, robust, and highly accurate counting of seated crowds.

  PublisherOA PDFDOI

• Comparative Analysis of Acted-Out and Real Patient Gait Parameters for Sensing Applications

  2025-03-17 · 1 citations

  articleSenior author

  Neurological conditions greatly affect patients’ mobility and can considerably shorten their lifespan. However, the shortage of neurologists can limit access to essential evaluations, particularly in underserved areas. On the other hand, alternative sensing methods (e.g., videos, WiFi, mmWave), based on machine learning, can provide cost-effective solutions, and have thus gained tractions in recent years. However, these systems need large/diverse datasets for development/evaluation, collection of which can be considerably challenging from real patients. This has then made recruiting actors to replicate gait disorders a common practice, raising questions about the similarity of their gait patterns to those of real patients and the generalizability of the subsequent findings. In this paper, we methodically investigate the extent to which actors can accurately replicate gait disorders, and the corresponding impact on gait disorder classification systems, marking the first study of its kind to the best of our knowledge. Specifically, we collect videos of real patients, largely based on collaboration with a neurology center, and further recruit actors to closely act them out. By statistically analyzing key gait parameters across the two datasets, we then reveal interesting underlying disparity patterns. We further develop gait disorder classifiers to illustrate the impact of this disparity on the classification performance. Our results show interesting trends, e.g., mild gait conditions are classified more accurately than what would have been with real patients, while moderate cases are classified less accurately. Overall, the paper provides a platform for interpreting the results when recruiting actors, thus paving the way towards equitable healthcare.

  PublisherDOI

Recent grants

• RI: Small: To Ask or Not to Ask - A Foundation for the Optimization of Human-Robot Networks

  NSF · $498k · 2016–2020

• NeTS: Small: Fundamentals of Assessing Occupancy Dynamics with Ubiquitous Wireless Signals

  NSF · $500k · 2018–2022

• RI-Small: An Integrative Framework for Communication and Motion-planning in Robotic Networks Operating in Fading Environments

  NSF · $350k · 2008–2012

• PECASE: Compressive Cooperative Sensing and Navigation in Mobile Networks

  NSF · $335k · 2012–2015

• CNS Core: Small: Fundamentals of Gait Disorder Assessment with Ubiquitous Wireless Signals

  NSF · $555k · 2022–2026

Frequent coauthors

• Miroslav Krstić

  University of California, San Diego

  35 shared

• Thomas Parisini

  34 shared

• João P. Hespanha

  34 shared

• Maria Elena Valcher

  31 shared

• Carolyn L. Beck

  31 shared

• Magnus Egerstedt

  31 shared

• Christos G. Cassandras

  29 shared

• Jing Sun

  Shaanxi University of Chinese Medicine

  27 shared